FRVT MORPH

Latest Report | Participation Agreement | API | Validation | Encryption

Overview

This page summarizes and links to all FRVT data and reports related to face morphing.
2023-03-06 NIST Interagency Report 8292 Draft Supplement: FRVT Part 4: MORPH - Performance of Automated Face Morph Detection PDF
2022-07-28 NIST Interagency Report 8430: FRVT Part 4A: Utility of 1:N Face Recognition Algorithms for Morph Detection PDF

[2023-03-06] An updated FRVT MORPH report has been published, adding results for one new algorithm submitted by West Virginia University (wvudiff-001).

Accuracy Summary

[last updated: 2023-03-06]

Morph Detection Performance

Morph Detection Performance

The table provides a summary of all algorithms measured on attack presentation classification error rate (APCER) when bona fide classification error rate (BPCER) is set to 0.1 and 0.01, across a subset of the different morphing datasets used in our evaluation. APCER, or morph miss rate, is the proportion of morphs that are incorrectly classified as bona fides (nonmorphs). BPCER, or false detection rate, is the proportion of bona fides falsely classified as morphs.

Face Recognition Accuracy on Morphs

Face Recognition Accuracy on Morphs

The graph below plots face recognition algorithm vulnerability on morphs against general algorithm accuracy on non-morphed photos. Each circular point represents a face recognition algorithm recently submitted to the NIST FRVT 1:1 activity, and each triangular point represents a face recognition algorithm submitted to the NIST FRVT MORPH activity. Note that algorithms submitted to FRVT 1:1 are not necessarily designed to handle morphed photos, and results are presented only as a point of reference. Submissions to FRVT MORPH are designed with goals of face recognition algorithm resistance against morphing. The y-axis plots MMPMR, which is the fraction of morphs where both subjects incorrectly match to the morph. The x-axis plots FNMR or miss rate on regular photos, which provides an indication of general algorithm accuracy. Both MMPMR and FNMR are calculated with thresholds set to where the false match rate (FMR) is 0.0001. The lower the MMPMR, the better the algorithm performs against morphs.
MMPMR vs. FNMR


The table below provides numerical tabulation of MMPMR and FNMR for recent face recognition algorithms submitted to FRVT MORPH and FRVT 1:1, ordered initially by FNMR.


Tier 1 - Low Quality Morphs

Website

Website


Global Morph

Global Morph


Tier 2 - Automated Morphs

Local Morph

Local Morph


Local Morph Colorized Average

Local Morph Colorized Average


Local Morph Colorized Match

Local Morph Colorized Match


UNIBO Automatic Morphed Face Generation Tool v1.0

UNIBO Automatic Morphed Face Generation Tool v1.0


DST

DST


Visa-Border

Visa-Border


UNIBO Automatic Morphed Face Generation Tool v2.0

UNIBO Automatic Morphed Face Generation Tool v2.0


Twente

Twente


MIPGAN-II

MIPGAN-II


Tier 3 - High Quality Morphs

Manual

Manual


Lincoln

Lincoln


Print and Scanned

Print and Scanned


Prior Editions of Report

All prior Ongoing FRVT MORPH reports can be accessed from here.

Overview

Face morphing and the ability to detect it is an area of high interest to a number of photo-credential issuance agencies and those employing face recognition for identity verification. The FRVT MORPH test will provide ongoing independent testing of prototype face morph detection technologies. The evaluation is designed to obtain an assessment on morph detection capability to inform developers and current and prospective end-users, and will evaluate two separate tasks:

  • Algorithmic capability to detect face morphing (morphed/blended faces) in still photographs
  • Face recognition algorithm resistance against morphing

Call for Data

We are seeking representative morph data to support our testing efforts. If your organization has morphs that 1) have never been shared with developers and 2) can be shared with NIST, please contact frvt@nist.gov.

How to Participate

To participate in this evaluation, developers need to submit a participation agreement to NIST, wrap their software behind the published C++ API, run their libraries through the provided validation package (which creates a submission package), encrypt the package, and provide a download link for the encrypted submission package. More details are provided below.

Participation Agreement

FRVT MORPH is conducted by NIST, an agency of the United States Government. Participation is free of charge. FRVT MORPH is open to a global audience of computer vision and face recognition developers. All organizations who seek to participate in FRVT MORPH must sign and submit all pages of this Participation Agreement. Note that this is a separate agreement from the FRVT Ongoing 1:1 agreement. [last update: 2018-06-11]

API Document

An updated version of the FRVT MORPH API document is now available. This update adds a second version of the detectMorphDifferentially() function that includes subject metadata as input to the algorithm. The additional subject metadata includes sex, age of the subject in the probe image, and the age/time difference between a suspected morph and the live probe image. [last update: 2022-05-19]

All FRVT APIs reference the supporting FRVT General Evaluation Specifications, which includes hardware and operating system environment, software requirements, reporting, and common data structures that support the APIs. Developers must ensure that their submission conforms to the API specifications.

Validation

An updated validation package has been published. All participants must run their software through the validation package prior to submission. The purpose of validation is to ensure consistent algorithm output between your execution and NIST’s execution. [last update: 2020-09-10]

Encryption

All submissions must be properly encrypted and signed before transmission to NIST. This must be done according to these instructions using the FRVT Ongoing public key linked from this page. Participants must email their public key to NIST. The participant’s public key must correspond to the participant’s public-key fingerprint provided on the signed Participation Application. [last update: 2018-05-10]

Submission

Encrypted files below 20MB can be emailed to NIST at frvt@nist.gov. Encrypted files above 20MB can be provided as a download link from a generic http webserver (e.g., Google Drive). We cannot accept Dropbox links. NIST will not register, or establish any kind of membership, on the provided website. Participants can submit their algorithm(s) as soon as the signed participation agreement is sent to NIST. There is no need to wait for NIST confirmation of the received agreement. [last update: 2018-05-10]

Interested Parties

Please contact NIST if

  • You are a developer of morph detection algorithms and/or a developer of face recognition algorithms that are resistant to face morphing
  • You represent an organization possessing suitable morph data that may be valuable to our effort
  • You have suggestions toward developing the technology or any more general interest in shaping how NIST proceeds with the FRVT MORPH program

Contact Information

Inquiries and comments may be submitted to frvt@nist.gov.

Subscribe to the FRVT mailing list to receive emails when announcements or updates are made.

Related NIST Projects

FRVT Ongoing
FRVT 1:1 Verification
FRVT 1:N Identification
FRVT Demographic Effects
FRVT Quality Assessment
FRVT Face Mask Effects
FRVT Paperless Travel
FRVT PAD

FRVT Twins Demonstration

Other Related Projects

FVC-onGoing Face Morphing Challenge